Dear Alex,
Did you follow the instructions for setting up the individual analyses along the lines
of the information shown in: http://www.fmrib.ox.ac.uk/fsl/feat5/perfusion.html ?
It sounds like your set up is exactly the same as this example.
Once you have this done and have specified the first-level contrast of interest,
all the subsequent modelling is the same as for any other FEAT analysis. You
just set up higher-level models to look at the difference of the first-level contrast
results between groups.
If you are concerned with your statistics being all zero, check the unthresholded
statistical images in the stats folder. This will show whether you have some
changes that are just below threshold or whether something has gone badly
wrong with your analysis and you are not generating sensible stats.
Hope this helps.
All the best,
Mark
On 29 Feb 2012, at 19:12, Alex Papoiu wrote:
> Hello everyone!
>
> My question relates to the proper setup for analysis in FSL to look at the baseline (resting/control) perfusion images for individual subjects? Subsequently, for doing a group analysis in 2 different patient groups (for example healthy controls and disease A) and third, for contrasting these two groups perfusion at rest? Importantly, we used ASL (arterial spin labeling).
>
> We tried to run the analysis as instructed in the FSL help files and the results were zero (at individual and also higher levels). We set-up one EV with one contrast with the value of 1. We are concerned whether we are doing this correctly and whether it has to be set-up differently for ASL data?
>
> Thank you kindly for your help. Alex
>
|